Comparing Annotated Datasets for Named Entity Recognition in English Literature